Junbo Wang
Papers
2
Total Citations
59
H-Index
2
About
Junbo Wang is a leading researcher in robotics, specializing in **robotic manipulation, imitation learning, and perception for articulated objects**. His work addresses the critical challenge of enabling robots to acquire diverse, generalizable skills for real-world, open-domain environments. Wang’s major contribution is advancing **one-shot imitation learning** and **robust perception under noisy conditions**, directly tackling the gap between controlled lab settings and unpredictable daily life. His highly cited paper, *"RH20T: A Comprehensive Robotic Dataset for Learning Diverse Skills in One-Shot"* (2024, 55 citations), provides a foundational dataset that has become a key resource for training robotic foundation models, significantly accelerating progress in skill transfer from a single demonstration. This work has already shaped how researchers approach generalizable robot learning. Additionally, his paper *"RPMArt: Towards Robust Perception and Manipulation for Articulated Objects"* (2024) pioneers methods to handle point cloud noise and perception failures, a persistent hurdle for manipulating common objects like doors and drawers. By focusing on both data infrastructure and algorithmic robustness, Wang’s research is pivotal for moving robots from factories into human-centric environments, making him a rising authority in practical, deployable robotic intelligence.
Research Focus
Key Achievements
Top Papers
- 1
- 2